On Wednesday, the hardest one day a week
Big middle, so hot today
May, 36-degree heat
Words drifting across the sky
The most handy room school pandas
Groupy DataFrame with Index Levels and Columns
It means hybrid packet and columns by index
Examples go on, (do not start writing examples, do not know how to explain it)
import pandas as pd
arrays = [['bar', 'bar', 'baz', 'baz', 'foo', 'foo', 'qux', 'qux'],
['one', 'two', 'one', 'two', 'one', 'two', 'one', 'two']]
index = pd.MultiIndex.from_arrays(arrays=arrays,names=['first','second'])
df = pd.DataFrame({'A':[3,1,4,5,9,2,6,1],
'B':[1,1,1,1,2,2,3,3]},index=index)
print(df)
复制代码There are examples, there are examples to show, right
A B
first second
bar one 3 1
two 1 1
baz one 4 1
two 5 1
foo one 9 2
two 2 2
qux one 6 3
two 1 3
复制代码Next, the big move to show the links
I want to be grouped by index index and second column B
Code is the first step, the effect will come later
grouped = df.groupby([pd.Grouper(level=1),'B']).sum()
print(grouped)
复制代码Note that there are two values groupby see, one pd.Grouper(level=1)the index of the second to the second B columns
The main is to make you understand, is how to calculate the packet oh ~
Of course, you can also be grouped by index name
df.groupby([pd.Grouper(level='second'), 'A']).sum()
复制代码And the above effect is the same as the
Even, we can directly abbreviated as
df.groupby(['second', 'A']).sum()
复制代码After the data packet portions can be selected, it can be an iterative
This section, in fact, we have achieved over
A chance to relive it
df = pd.DataFrame({'A':['bar', 'bar', 'foo', 'foo', 'foo', 'foo', 'foo'],
'B':['one', 'two', 'one', 'two', 'one', 'two', 'three'],
'C':[3,1,4,5,9,2,6],
'D':[1,1,1,1,2,2,3]})
print(df)
grouped = df.groupby('A')
for name,group in grouped:
print(name)
print(group)
复制代码See the name of the grouping are bar and foo, familiar with it, normal operation
Iteration time, loop can be used for in
bar
A B C D
0 bar one 3 1
1 bar two 1 1
foo
A B C D
2 foo one 4 1
3 foo two 5 1
4 foo one 9 2
5 foo two 2 2
6 foo three 6 3
复制代码If multiple grouped keys, e.g.groupby(['A','B'])
It will naturally form a tuple name
bars = grouped.get_group('bar') # 通过分组的名字
print(bars)
复制代码And the other one?
df.groupby(['A', 'B']).get_group(('bar', 'one'))
复制代码Well, for myself, so I write, it is more of a
A large degree of difficulty, coming, aggregate functions
First look at the built-in aggregate functions
sum(), mean(), max(), min(), count(), size(), describe()
复制代码Actually only so few, it is because I did not write the whole
We have been operating this many times
Then you can look at a more advanced level,
可自定义函数，传入agg方法中
复制代码We just analyzed by data
A B C D
0 bar one 3 1
1 bar two 1 1
2 foo one 4 1
3 foo two 5 1
4 foo one 9 2
5 foo two 2 2
6 foo three 6 3
复制代码A and B in accordance with the packet has two values A, B has three values, five groups formed after the packets
Look, do not blink, the operation
grouped = df.groupby(['A','B'])
print(grouped.agg('mean'))
复制代码grouped = df.groupby(['A','B'])
print(grouped['C'].agg('mean'))
复制代码Continuing conversion idea, a plurality of separate aggregate functions
print(grouped['C'].agg(['mean','sum']))
复制代码Very powerful, learned it
Continue to be, do not be afraid, find a variety of aggregation operation at the same time change the column name
print(grouped['C'].agg([('A','mean'),('B','max')]))
复制代码Different columns to use aggregation functions
print(grouped.agg({'C':['sum','mean'],'D':['min','max']}))
复制代码These are agg dry, I can continue to compile oh ~
groupby, the index may be modified to form non-core noted a parameter addedas_index=False
grouped = df.groupby(['A','B'],as_index=False)
print(grouped.agg({'C':['sum','mean'],'D':['min','max']}))
复制代码The last operation, agg aggregate functions which can be customized
Generally, this is the case, I do, of course, is no exception friends
grouped = df.groupby('A')
def max_min(group):
return group.max()-group.min()
print(grouped.agg(max_min))
复制代码AGG (custom function)
This place custom functions, also supports lambdaoh ~
He confused, confused all right, take live phone line
Shot here, in this shot
Reproduced in: https: //juejin.im/post/5d088203f265da1bd260edc4